Clustering sustainable tourism destinations through Instagram photo analysis: a machine learning approach

Purpose This study uses advanced machine learning (ML) techniques to analyze Instagram photos and cluster sustainable tourism destinations. By revealing visual patterns and themes, it provides valuable insights into user-generated content (UGC) and its influence on destination image formation. Using established methods such as k-means clustering and semantic analysis, this study aims to enhance sustainable tourism promotion by optimizing targeted marketing strategies and improving the overall tourist experience. Design/methodology/approach This study uses ML to analyze Instagram images through a four-step approach. The authors collected 208 posts and 1,890 images, ensuring data accuracy via cleansing. Each image was labeled for machine understanding using the Doc2Vec model to capture semantic meaning from captions. Finally, the authors applied k-means clustering to group posts based on similarities in visual motifs and hashtag usage, enhancing interpretability and contributing to the field of image classification through advanced ML techniques. Findings This study demonstrates that ML effectively clusters sustainable tourism destinations on Instagram, identifying four key themes: “Natural landscapes,” “Historical and archaeological,” “Branding and graphics” and “Cultural heritage.” Popular destinations like Marrakech, Aït Bouguemez and Dakhla dominate image concentrations. These insights provide actionable strategies for promoting sustainability, enhancing destination branding and tailoring marketing approaches for Morocco and similar global destinations. Practical implications The research contributes to the global effort to address how Morocco’s commitment to sustainable tourism can be included in the Sustainable Development Goals. It provides insights into forecasting regional tourism demand and coastal management strategies for sustainable tourism development in Morocco. Originality/value This paper offers data-driven insights into the visual representation of sustainable tourism destinations, thereby providing a valuable resource for policymakers and practitioners seeking to inform their decisions and actions. The proposed method can be leveraged to inform policy and practice in sustainable tourism, ultimately supporting local businesses and promoting eco-friendly travel practices.